BOOK

ggpairs(col_I_mixed[,c("log-ADD_FEE", "BOOK", "PPIND")], aes(alpha = 0.5, color=PPIND),
        diag = list(continuous = wrap("barDiag", bins = 20)),
        upper = list(continuous = wrap("cor", size = 2)))

## Warning: Removed 40 rows containing non-finite outside the scale range
## (`stat_bin()`).

## Warning in ggally_statistic(data = data, mapping = mapping, na.rm = na.rm, :
## Removed 40 rows containing missing values

## Warning: Removed 40 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

## Warning: Removed 40 rows containing missing values or values outside the scale range
## (`geom_point()`).

## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_bin()`).

## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 40 rows containing non-finite outside the scale range
## (`stat_bin()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_bin()`).

col_I_sn|>filter(BOOK>=900 | BOOK <= 350)

## # A tibble: 4 × 49
##   ...1     PPIND  FICE STATE TYPE  AVRMATH AVRVERB AVRCOMB AVR_ACT MATH_1 MATH_3
##   <chr>    <chr> <dbl> <chr> <chr>   <dbl>   <dbl>   <dbl>   <dbl>  <dbl>  <dbl>
## 1 Howard … priv…  1448 DC    I         457     426     883      20    390    510
## 2 Tulane … priv…  2029 LA    I          NA      NA      NA      NA    560    670
## 3 Hofstra… priv…  2732 NY    I         540     470    1010      24    480    590
## 4 Renssel… priv…  2803 NY    I          NA      NA      NA      28    610    710
## # ℹ 38 more variables: VERB_1 <dbl>, VERB_3 <dbl>, ACT_1 <dbl>, ACT_3 <dbl>,
## #   APP_REC <dbl>, APP_ACC <dbl>, NEW_STUD <dbl>, NEW10 <dbl>, NEW25 <dbl>,
## #   FULLTIME <dbl>, PARTTIME <dbl>, IN_STATE <dbl>, OUT_STAT <dbl>,
## #   R_B_COST <dbl>, ROOM <dbl>, BOARD <dbl>, ADD_FEE <dbl>, BOOK <dbl>,
## #   PERSONAL <dbl>, PH_D <dbl>, TERM_D <dbl>, SF_RATIO <dbl>, DONATE <dbl>,
## #   INSTRUCT <dbl>, GRADUAT <dbl>, SAL_FULL <dbl>, SAL_AC <dbl>, SAL_AS <dbl>,
## #   SAL_ALL <dbl>, COMP_FUL <dbl>, COMP_AC <dbl>, COMP_AS <dbl>, …

Может быть как низкой, так и высокой в зависимости от штата, политики университета и скрытых факторов.

ADD_FEE

col_I_sn|>filter(ADD_FEE>3000)

## # A tibble: 7 × 49
##   ...1     PPIND  FICE STATE TYPE  AVRMATH AVRVERB AVRCOMB AVR_ACT MATH_1 MATH_3
##   <chr>    <chr> <dbl> <chr> <chr>   <dbl>   <dbl>   <dbl>   <dbl>  <dbl>  <dbl>
## 1 Univers… publ…  1313 CA    I         585     485    1070      24    520    650
## 2 Univers… publ…  1315 CA    I         621     520    1141      NA    558    700
## 3 Univers… publ…  1316 CA    I         542     440     982      23    470    620
## 4 Univers… publ…  1317 CA    I         616     515    1131      26    560    680
## 5 Univers… publ…  1320 CA    I         547     459    1006      NA    520    610
## 6 Univers… publ…  1321 CA    I         554     505    1059      NA    490    620
## 7 Univers… publ…  2221 MA    I          NA      NA      NA      NA    460    590
## # ℹ 38 more variables: VERB_1 <dbl>, VERB_3 <dbl>, ACT_1 <dbl>, ACT_3 <dbl>,
## #   APP_REC <dbl>, APP_ACC <dbl>, NEW_STUD <dbl>, NEW10 <dbl>, NEW25 <dbl>,
## #   FULLTIME <dbl>, PARTTIME <dbl>, IN_STATE <dbl>, OUT_STAT <dbl>,
## #   R_B_COST <dbl>, ROOM <dbl>, BOARD <dbl>, ADD_FEE <dbl>, BOOK <dbl>,
## #   PERSONAL <dbl>, PH_D <dbl>, TERM_D <dbl>, SF_RATIO <dbl>, DONATE <dbl>,
## #   INSTRUCT <dbl>, GRADUAT <dbl>, SAL_FULL <dbl>, SAL_AC <dbl>, SAL_AS <dbl>,
## #   SAL_ALL <dbl>, COMP_FUL <dbl>, COMP_AC <dbl>, COMP_AS <dbl>, …

University of California, Massachusetts - особые индивиды (из-за ВВП).

AVRCOMB/NEW10

ggpairs(col_I_mixed[colNew], aes(alpha = 0.5, color=PPIND),
        diag = list(continuous = wrap("barDiag", bins = 20)),
        upper = list(continuous = wrap("cor", size = 2)))

## Warning: Removed 16 rows containing non-finite outside the scale range
## (`stat_bin()`).

## Warning in ggally_statistic(data = data, mapping = mapping, na.rm = na.rm, :
## Removed 75 rows containing missing values

## Warning: Removed 16 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

## Warning: Removed 75 rows containing missing values or values outside the scale range
## (`geom_point()`).

## Warning: Removed 67 rows containing non-finite outside the scale range
## (`stat_bin()`).

## Warning: Removed 67 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 16 rows containing non-finite outside the scale range
## (`stat_bin()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 67 rows containing non-finite outside the scale range
## (`stat_bin()`).

col_I_sn|>filter(AVRCOMB<=1200 & NEW10>=75)

## # A tibble: 8 × 49
##   ...1     PPIND  FICE STATE TYPE  AVRMATH AVRVERB AVRCOMB AVR_ACT MATH_1 MATH_3
##   <chr>    <chr> <dbl> <chr> <chr>   <dbl>   <dbl>   <dbl>   <dbl>  <dbl>  <dbl>
## 1 Univers… publ…  1313 CA    I         585     485    1070      24    520    650
## 2 Univers… publ…  1314 CA    I         574     455    1029      NA    510    640
## 3 Univers… publ…  1315 CA    I         621     520    1141      NA    558    700
## 4 Univers… publ…  1316 CA    I         542     440     982      23    470    620
## 5 Univers… publ…  1317 CA    I         616     515    1131      26    560    680
## 6 Univers… publ…  1320 CA    I         547     459    1006      NA    520    610
## 7 Univers… publ…  1321 CA    I         554     505    1059      NA    490    620
## 8 Univers… publ…  2974 NC    I         594     527    1121      NA    530    530
## # ℹ 38 more variables: VERB_1 <dbl>, VERB_3 <dbl>, ACT_1 <dbl>, ACT_3 <dbl>,
## #   APP_REC <dbl>, APP_ACC <dbl>, NEW_STUD <dbl>, NEW10 <dbl>, NEW25 <dbl>,
## #   FULLTIME <dbl>, PARTTIME <dbl>, IN_STATE <dbl>, OUT_STAT <dbl>,
## #   R_B_COST <dbl>, ROOM <dbl>, BOARD <dbl>, ADD_FEE <dbl>, BOOK <dbl>,
## #   PERSONAL <dbl>, PH_D <dbl>, TERM_D <dbl>, SF_RATIO <dbl>, DONATE <dbl>,
## #   INSTRUCT <dbl>, GRADUAT <dbl>, SAL_FULL <dbl>, SAL_AC <dbl>, SAL_AS <dbl>,
## #   SAL_ALL <dbl>, COMP_FUL <dbl>, COMP_AC <dbl>, COMP_AS <dbl>, …

University of California, North Carolina. Низкий порог для поступления?

PH_D

ggpairs(col_I_mixed[colStud], aes(alpha = 0.5, color=PPIND),
        diag = list(continuous = wrap("barDiag", bins = 20)),
        upper = list(continuous = wrap("cor", size = 2)))

## Warning in ggally_statistic(data = data, mapping = mapping, na.rm = na.rm, :
## Removed 9 rows containing missing values

## Warning in ggally_statistic(data = data, mapping = mapping, na.rm = na.rm, :
## Removed 5 rows containing missing values

## Warning: Removed 9 rows containing missing values or values outside the scale range
## (`geom_point()`).

## Warning: Removed 9 rows containing non-finite outside the scale range
## (`stat_bin()`).

## Warning in ggally_statistic(data = data, mapping = mapping, na.rm = na.rm, :
## Removed 14 rows containing missing values

## Warning: Removed 9 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

## Warning: Removed 5 rows containing missing values or values outside the scale range
## (`geom_point()`).

## Warning: Removed 14 rows containing missing values or values outside the scale range
## (`geom_point()`).

## Warning: Removed 5 rows containing non-finite outside the scale range
## (`stat_bin()`).

## Warning: Removed 5 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 9 rows containing non-finite outside the scale range
## (`stat_bin()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 5 rows containing non-finite outside the scale range
## (`stat_bin()`).

col_I_sn|>filter(PH_D<=70)

## # A tibble: 5 × 49
##   ...1     PPIND  FICE STATE TYPE  AVRMATH AVRVERB AVRCOMB AVR_ACT MATH_1 MATH_3
##   <chr>    <chr> <dbl> <chr> <chr>   <dbl>   <dbl>   <dbl>   <dbl>  <dbl>  <dbl>
## 1 Univers… publ…  1063 AK    I         499     462     961      22     NA     NA
## 2 Ball St… publ…  1786 IN    I         473     422     895      21    400    530
## 3 Univers… publ…  2031 LA    I          NA      NA      NA      19     NA     NA
## 4 Texas S… publ…  3642 TX    I         390     420     810      19     NA     NA
## 5 Texas W… publ…  3646 TX    I          NA      NA      NA      NA    350    480
## # ℹ 38 more variables: VERB_1 <dbl>, VERB_3 <dbl>, ACT_1 <dbl>, ACT_3 <dbl>,
## #   APP_REC <dbl>, APP_ACC <dbl>, NEW_STUD <dbl>, NEW10 <dbl>, NEW25 <dbl>,
## #   FULLTIME <dbl>, PARTTIME <dbl>, IN_STATE <dbl>, OUT_STAT <dbl>,
## #   R_B_COST <dbl>, ROOM <dbl>, BOARD <dbl>, ADD_FEE <dbl>, BOOK <dbl>,
## #   PERSONAL <dbl>, PH_D <dbl>, TERM_D <dbl>, SF_RATIO <dbl>, DONATE <dbl>,
## #   INSTRUCT <dbl>, GRADUAT <dbl>, SAL_FULL <dbl>, SAL_AC <dbl>, SAL_AS <dbl>,
## #   SAL_ALL <dbl>, COMP_FUL <dbl>, COMP_AC <dbl>, COMP_AS <dbl>, …

Незначительные выбросы - особые индивиды. Могут быть связаны с низким GRADUAT и высоким SF_RATIO.

OUT_STAT

ggpairs(col_I_mixed[,c("R_B_COST", "OUT_STAT", "PPIND")], aes(alpha = 0.5, color=PPIND),
        diag = list(continuous = wrap("barDiag", bins = 20)),
        upper = list(continuous = wrap("cor", size = 2)))

## Warning in ggally_statistic(data = data, mapping = mapping, na.rm = na.rm, :
## Removing 1 row that contained a missing value

## Warning: Removed 1 row containing missing values or values outside the scale range
## (`geom_point()`).

## Warning: Removed 1 row containing non-finite outside the scale range
## (`stat_bin()`).

## Warning: Removed 1 row containing non-finite outside the scale range
## (`stat_boxplot()`).

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## Warning: Removed 1 row containing non-finite outside the scale range
## (`stat_bin()`).

col_I_sn|>filter(OUT_STAT>=14000 & PPIND=="public")

## # A tibble: 2 × 49
##   ...1     PPIND  FICE STATE TYPE  AVRMATH AVRVERB AVRCOMB AVR_ACT MATH_1 MATH_3
##   <chr>    <chr> <dbl> <chr> <chr>   <dbl>   <dbl>   <dbl>   <dbl>  <dbl>  <dbl>
## 1 Univers… publ…  3696 VT    I         553     488    1041      NA    500    610
## 2 Univers… publ…  9092 MI    I         634     543    1177      27    580    700
## # ℹ 38 more variables: VERB_1 <dbl>, VERB_3 <dbl>, ACT_1 <dbl>, ACT_3 <dbl>,
## #   APP_REC <dbl>, APP_ACC <dbl>, NEW_STUD <dbl>, NEW10 <dbl>, NEW25 <dbl>,
## #   FULLTIME <dbl>, PARTTIME <dbl>, IN_STATE <dbl>, OUT_STAT <dbl>,
## #   R_B_COST <dbl>, ROOM <dbl>, BOARD <dbl>, ADD_FEE <dbl>, BOOK <dbl>,
## #   PERSONAL <dbl>, PH_D <dbl>, TERM_D <dbl>, SF_RATIO <dbl>, DONATE <dbl>,
## #   INSTRUCT <dbl>, GRADUAT <dbl>, SAL_FULL <dbl>, SAL_AC <dbl>, SAL_AS <dbl>,
## #   SAL_ALL <dbl>, COMP_FUL <dbl>, COMP_AC <dbl>, COMP_AS <dbl>, …

col_I_sn|>filter(OUT_STAT<=10500 & PPIND=="private")

## # A tibble: 6 × 49
##   ...1     PPIND  FICE STATE TYPE  AVRMATH AVRVERB AVRCOMB AVR_ACT MATH_1 MATH_3
##   <chr>    <chr> <dbl> <chr> <chr>   <dbl>   <dbl>   <dbl>   <dbl>  <dbl>  <dbl>
## 1 Univers… priv…  1431 DE    I          NA      NA    1035      NA    500    610
## 2 Howard … priv…  1448 DC    I         457     426     883      20    390    510
## 3 Rice Un… priv…  3604 TX    I          NA      NA      NA      NA    665    763
## 4 Texas C… priv…  3636 TX    I          NA      NA      NA      NA    460    600
## 5 Brigham… priv…  3670 UT    I          NA      NA      NA      NA     NA     NA
## 6 Baylor … priv…  6967 TX    I          NA      NA      NA      NA    500    620
## # ℹ 38 more variables: VERB_1 <dbl>, VERB_3 <dbl>, ACT_1 <dbl>, ACT_3 <dbl>,
## #   APP_REC <dbl>, APP_ACC <dbl>, NEW_STUD <dbl>, NEW10 <dbl>, NEW25 <dbl>,
## #   FULLTIME <dbl>, PARTTIME <dbl>, IN_STATE <dbl>, OUT_STAT <dbl>,
## #   R_B_COST <dbl>, ROOM <dbl>, BOARD <dbl>, ADD_FEE <dbl>, BOOK <dbl>,
## #   PERSONAL <dbl>, PH_D <dbl>, TERM_D <dbl>, SF_RATIO <dbl>, DONATE <dbl>,
## #   INSTRUCT <dbl>, GRADUAT <dbl>, SAL_FULL <dbl>, SAL_AC <dbl>, SAL_AS <dbl>,
## #   SAL_ALL <dbl>, COMP_FUL <dbl>, COMP_AC <dbl>, COMP_AS <dbl>, …

Цены на обучение варьируются по штатам и самим университетам.

Public: проверка на нормальность, тесты

# colSums(is.na(col_split_mixed$public))
col_norm_public<-col_split_mixed$public|>perform_normality_tests(colMixed)

## Warning in FUN(X[[i]], ...): Removed 33 NA values from column 'log-ADD_FEE'
## before testing.

## Warning in FUN(X[[i]], ...): Removed 2 NA values from column 'BOOK' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 1 NA values from column 'OUT_STAT' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 15 NA values from column 'NEW10' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 43 NA values from column 'AVRCOMB' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 5 NA values from column 'PH_D' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 4 NA values from column 'GRADUAT' before
## testing.

col_norm_public$plots

## $`log-ADD_FEE`

## 
## $BOOK

## 
## $R_B_COST

## 
## $OUT_STAT

## 
## $NEW10

## 
## $AVRCOMB

## 
## $SF_RATIO

## 
## $PH_D

## 
## $GRADUAT

col_norm_public$results|>dplyr::select(-Statistic)|>rmarkdown::paged_table()

Private: проверка на нормальность, тесты

col_norm_private<-col_split_mixed$private|>perform_normality_tests(colMixed)

## Warning in FUN(X[[i]], ...): Removed 7 NA values from column 'log-ADD_FEE'
## before testing.

## Warning in FUN(X[[i]], ...): Removed 1 NA values from column 'NEW10' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 24 NA values from column 'AVRCOMB' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 4 NA values from column 'PH_D' before
## testing.

## Warning in FUN(X[[i]], ...): Removed 1 NA values from column 'GRADUAT' before
## testing.

col_norm_private$plots

## $`log-ADD_FEE`

## 
## $BOOK

## 
## $R_B_COST

## 
## $OUT_STAT

## 
## $NEW10

## 
## $AVRCOMB

## 
## $SF_RATIO

## 
## $PH_D

## 
## $GRADUAT

col_norm_private$results|>dplyr::select(-Statistic)|>rmarkdown::paged_table()